Skip to main content

Topic: denon alpha processing (Read 6529 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • cool n hot
  • [*]
denon alpha processing
Hi. I've been reading info about how denon's alpha processing works and have become confused about what it actually does. Is it a sort of post dithering and or is it used to improve the accuracy of pre dithered sources or is it something different altogether. Please help me out . Thank you.

  • _if
  • [*][*]
denon alpha processing
Reply #1
Unless their Glossary of Technical Terms page is giving an inaccurate description of what it does, it cannot be anything more than a weak attempt at marketing hype. The description contains multiple falsehoods about how digital audio works while touting their processor's features as if they're useful or unique. Adding 16 bits to an already 16 bit audio stream will do literally nothing. They're empty 0s, no data in them, least of all data that was contained in the original sound before it was recorded. If, for some reason, they're not empty 0s, whatever is contained in those extra 16 bits is not supposed to be there and should probably be counted as noise or distortion. The upsampling feature is stock on virtually all digital to analog converters at any price point. They say theirs does 16 times 44.1 kHz; I don't know exactly what rate DACs usually use, hopefully someone more knowledgeable in that area can chime in.

  • AndyH-ha
  • [*][*][*][*][*]
denon alpha processing
Reply #2
64X and 128X are very commonly used factors.

  • bennetng
  • [*][*][*][*][*]
denon alpha processing
Reply #3
RMAA result of a $6xx Denon CD player (DN-C635)
http://www.hydrogenaudio.org/forums/index....ost&id=7792

Played through balanced XLR output and recorded by RME Multiface II. SPDIF result also included to show that the disc is not defective or incompatible with the player.
While $6xx for a standalone CD player is not ridiculously expensive, the result is somewhat disappointing.

http://www.denon.com/pages/GlossaryDetail.aspx?GId=15
Then why there are so many sidebands in the RMAA result? 

The 7z archive posted above also include RMAA result of my PC's onboard VIA HD audio interface recorded by X-Fi Titanium HD for comparison.

  • saratoga
  • [*][*][*][*][*]
denon alpha processing
Reply #4
64X and 128X are very commonly used factors.


Or even higher.  The Sansa Clip players operate at 768x for 44.1k IIRC.

  • DVDdoug
  • [*][*][*][*][*]
denon alpha processing
Reply #5
Quote
While $6xx for a standalone CD player is not ridiculously expensive, the result is somewhat disappointing.
$600 USD sounds like a LOT to me,.  For that price I could buy 10 (or more) CD players or 5 (or more) Blu-Ray players that also play CDs & DVDs, all with sound quality that's better than human hearing.

Now, if you are into "build quality" or you like the "look and feel", or the brand name, etc., it might be worth it to you.

Personally, I get annoyed if I have something expensive and it breaks down, because the repair cost is more than replacing it with something cheaper...    Sometimes a repair costs more than replacing it with an identical/similar unit.    If my "Wal-Mart Special" dies, I can just run over to Wal-Mart and pick-up another one and I don't feel too bad about it. 



----------
For about the same price range....  I had to shop-around for it, but I bought a laptop for around $700 that plays (and records) all kinds of discs, including Blu-Ray, and it it has an HDMI port for connection to my home theater system.  Plus, my entire music collection of about 15,000 MP3s is on the hard drive.

I'm not saying you should replace your CD player with a computer, or use MP3s instead of CDs...  I'm just saying you are not getting much bang-for- the-buck with a $600 CD player.

denon alpha processing
Reply #6
Hi. I've been reading info about how denon's alpha processing works and have become confused about what it actually does. Is it a sort of post dithering and or is it used to improve the accuracy of pre dithered sources or is it something different altogether. Please help me out . Thank you.


Their first illustation is a false claim:

Denon's false claim:

http://www.denon.com/pages/GlossaryDetail.aspx?GId=13



The real truth:

http://www.stereophile.com/content/philips...er-measurements



-90 dB FS dithered sine wave
  • Last Edit: 27 January, 2014, 03:41:30 PM by Arnold B. Krueger

  • WernerO
  • [*][*]
denon alpha processing
Reply #7
The upsampling feature ... They say theirs does 16 times 44.1 kHz; I don't know exactly what rate DACs usually use, hopefully someone more knowledgeable in that area can chime in.


In oversampling there are two factors of relevance.

The first is the amount of oversampling encompassing the DAC's recontruction filtering. This is typically 8x, perhaps 16x (as Denon claim), the second factor is the increase in sample rate on the path from converting long-word PCM to the short-word stream at the DAC's delta-sigma modulator. The product of the two gives the high 256, 384, etc... seen in the blurbs.


  • knutinh
  • [*][*][*][*][*]
denon alpha processing
Reply #8
...Adding 16 bits to an already 16 bit audio stream will do literally nothing. They're empty 0s, no data in them, least of all data that was contained in the original sound before it was recorded. If, for some reason, they're not empty 0s, whatever is contained in those extra 16 bits is not supposed to be there and should probably be counted as noise or distortion...

If you are upsampling a signal, you are (in principle) inserting zeros samples and lowpass filtering. This filtering might introduce new samples that falls "in-between" the levels of a 16-bit input, and would either have to be rounded/dithered to 16 bits or to a higher resolution output (e.g. 24 bits).

Of course, marketing might misinterpret this as "adding more bits to the signal" when it is rather a more mundane "converting the signal to a continous waveform".

-k
  • Last Edit: 28 January, 2014, 02:39:30 AM by knutinh

  • bennetng
  • [*][*][*][*][*]
denon alpha processing
Reply #9
Quote
While $6xx for a standalone CD player is not ridiculously expensive, the result is somewhat disappointing.
$600 USD sounds like a LOT to me,.  For that price I could buy 10 (or more) CD players or 5 (or more) Blu-Ray players that also play CDs & DVDs, all with sound quality that's better than human hearing.

I'm not saying you should replace your CD player with a computer, or use MP3s instead of CDs...  I'm just saying you are not getting much bang-for- the-buck with a $600 CD player.

So true. In fact that CD player isn't mine. 
I am pretty sure that someone will recommend $6000 USD CD transports and DACs if I post in some other forums.

  • _if
  • [*][*]
denon alpha processing
Reply #10
...Adding 16 bits to an already 16 bit audio stream will do literally nothing. They're empty 0s, no data in them, least of all data that was contained in the original sound before it was recorded. If, for some reason, they're not empty 0s, whatever is contained in those extra 16 bits is not supposed to be there and should probably be counted as noise or distortion...

If you are upsampling a signal, you are (in principle) inserting zeros samples and lowpass filtering. This filtering might introduce new samples that falls "in-between" the levels of a 16-bit input, and would either have to be rounded/dithered to 16 bits or to a higher resolution output (e.g. 24 bits).

Of course, marketing might misinterpret this as "adding more bits to the signal" when it is rather a more mundane "converting the signal to a continous waveform".

-k

Right, so all this adds up to is "we're doing something pretty much routine but describing it in more detail than you usually hear, making it seem like it's really special." Unless they "process" the signal before it reaches the DAC (I'm unclear if ALPHA Processing is a DSP or a description of the normal functioning of a DAC), in which case, correct me if I'm wrong, it would be as I had said originally and doing nothing useful.

denon alpha processing
Reply #11
I'm unclear if ALPHA Processing is a DSP or a description of the normal functioning of a DAC
Their impulse response pictures look rather minimum-phase, not linear-phase like most modern DACs. At the same time they claim less ripple and no aliasing. Usually this comes at a price so it would be interesting to see all aspects of the process, but that's probably not something to be used in marketing

  • bennetng
  • [*][*][*][*][*]
denon alpha processing
Reply #12
Usually this comes at a price

Such as high fequency rolloff?

Also that -90dB sine wave looks impressive. I would like to see how Denon process a complex signal such as typical music files instead of test signals. Just hope it is not done by converting a high amplitude 16-bit signal to 32-bit then attenuate to -90dB.

denon alpha processing
Reply #13
Usually this comes at a price
Such as high fequency rolloff?
Yes. No aliasing means that the transition band has to end at the Nyquist frequency. To keep the transition band short a steep filter has to be used. With a more gentle filter one has to sacrifice the upper part of the pass band (HF roll off).
Also that -90dB sine wave looks impressive.
They state that this signal comes from a cd, so 16/44.1. I've never seen such a clean -90dBFS 1kHz sine that was properly 16-bit dithered, unless you filter out all other frequencies. But that wouldn't work for broadband signals like music
OTOH that staircase "sine" doesn't look like a properly dithered signal, so I remain confused about what problem Denon is trying to solve.
  • Last Edit: 31 January, 2014, 04:00:17 AM by Kees de Visser

  • probedb
  • [*][*][*][*][*]
denon alpha processing
Reply #14
Remember the stuff on Denon's website will have been done for marketing purposes by people in marketing, not for anything technical. Just to confuse your average audiophile so they can sell stuff